Expression of normally silent genes
The dysregulated uncontrolled cell division of the cancer cell creates a
milieu in which the products of normally silent genes may be expressed.
Sometimes these encode differentiation antigens normally associated with
an earlier developmental stage. Thus tumors derived from the same cell
type are often found to express such oncofetal antigens that are also
present on embryonic cells. Examples would be α-fetoprotein in hepatic
carcinoma and carcinoembryonic antigen (CEA) in cancer of the intestine.
Certain monoclonal antibodies also react with tumors of neural crest origin
and fetal melanocytes. Another monoclonal antibody defines the SSEA-1
antigen found on a variety of human tumors and early mouse embryos but
absent from adult cells with the exception of human granulocytes and
monocytes.
But the exciting quantum leap forward stems from the original
observation that cytosolic viral nucleoprotein could provide a target for Tc
cells by appearing on the cell surface as a processed peptide associated with
MHC class I (cf. p. 122). This established the general principle that the
intracellular proteins that are not destined to be positioned in the surface
plasma membrane can still signal their presence to T-cells in the outer world
by the processed peptide-MHC mechanism. Cytotoxic T-cells specific for
tumor cells, obtained from mixed cultures of peripheral blood cells with
tumor, can be used to establish the identity of the antigen employing the
strategy described in Figure 17.7. By something of a tour deforce a gene
encoding a melanoma antigen, MAGE-1, was identified. It belongs to a
family of 12 genes, 6 of which are expressed in a significant proportion of
melanomas as well as head and neck tumors, nonsmall cell lung cancers and
bladder carcinomas. MAGE-1 is not expressed in normal tissues except for
germline cells in testis and gives rise to antigenic T-cell epitopes that, in the
light of the absence of class I MHC on the testis cells, must be considered
tumor-specific. This exciting research reveals the tumor-specific antigen as
an expression of a normally silent gene.
Mutant antigens
The seminal work on tummutants (Milestone 17.1) has persuaded us that
single point mutations in oncogenes can account for the large diversity of
antigens found on carcinogen-induced tumors. The specific immunity
provoked by chemically induced tumors can be elicited by heat-shock
protein 70 (hsp70) and hsp90 isolated from the tumor cells, but their
immunogenicity is lost when the associated low-molecular-weight peptides
are removed. These peptides could, however, stimulate the specific CD8
cytotoxic T-cell clones generated by the tumors, and three possible
mechanisms have been advanced to account for the enhancement of tumor
immune responses by hsps. First, they can act as “danger” signals by
activating antigen-presenting cells. Second, necrotic tumor cells expressing
the hsps can transfer hsp-peptide complexes to host antigen-presenting cells
where they can cross-prime cytotoxic CD8 T-cells through the MHC class I
endogenous presentation pathway. And last, the hsps may influence the
capacity of the tumor cell itself to process and present endogenous mutated
and, of course, “silent” antigens as targets for specific T-cells (Figure 17.9).
There is considerable evidence for the production of mutated peptides in
human tumors. The gene encoding cell cycle checkpoint protein, p53, is a
hotspot for mutation in numerous cancers. The mutant forms of p53 that are
frequently found in tumors represent loss-of-function mutants that fail to
arrest division of cells that have suffered DNA damage; such damage would
normally trigger cell cycle arrest or apoptosis of the afflicted cell. The
oncogenic human ras genes differ from their normal counterparts by point
mutations usually leading to single amino acid substitutions in positions 12,
13 or 61. Such mutations generate constitutively active forms of Ras that
promote increased rates of cell division through activation of the MAPK
pathway (see p. 211), and have been recorded in 40% of human colorectal
cancers and in more than 90% of pancreatic carcinomas, as well as other
malignancies. The mutated ras peptide can induce proliferative T-cell lines
in vitro.
Figure 17.9. The role of heat-shock proteins (hsps) in tumor
immunogenicity.
(1) Stress factors upregulate hsps that can form complexes with processed
tumor antigen (Ag) and increase surface presentation of antigenic peptide
by MHC class I. (2) They can also lead to necrosis and release of hsp-
peptide complexes, which (3) can act as stimulatory danger signals to
dendritic antigen-presenting cells and penetrate the cytoplasm, where (4)
they can enter the MHC class I processing pathway by so-called cross-
priming. (5) CD8 resting T-cells become activated and (6) kill the tumor
cells. CTL, cytotoxic T-lymphocytes. (Based on Wells A.D. & Malkowsky
M. (2000) Immunology Today 21 , 129.)
Changes in carbohydrate structure
The chaotic internal control of metabolism within neoplastic cells often
leads to the presentation of abnormal carbohydrate structures on the cell
surface. Sometimes one sees blocked synthesis, e.g. deletion of blood group
A. In other cases there may be enhanced synthesis of structures absent in
progenitor cells: thus some gastrointestinal cancers express the Lewis Le a
−
−
antigen in individuals who are Le(a , b ) and others produce extended
a
chains bearing dimeric Le or Le(a,b).
Abnormal mucin synthesis can have immunological consequences.
Consider the mucins of pancreatic and breast tissue. These consist of a
polypeptide core of 20-amino acid tandem repeats with truly abundant O-
linked carbohydrate chains. A monoclonal antibody SM-3 directed to the
core polypeptide reacts poorly with normal tissue where the epitope is
masked by glycosylation, but well with breast and pancreatic carcinomas
possessing shorter and fewer O-linked chains. Tc cells specific for tumor
mucins are not MHC restricted and the slightly heretical suggestion has
been made that the T-cell receptors (TCRs) are binding multivalently to
closely spaced SM-3 epitopes on unprocessed mucins; alternatively, and
closer to the party line, recognition is by γδ cells.
Molecules related to metastatic potential
Changes in surface carbohydrates can have a dramatic effect on
x
malignancy. For example, colonic cancers expressing sialyl Le have a poor
prognosis and higher propensity to metastasize. Lung cancer patients whose
tumors showed deletion of blood group A had a much worse prognosis than
y
those with continuous A; the finding that patients expressing H/Le /Le b
also had a poorer prognosis than antigen-negative subjects is consistent
with this observation.
The role of CD44 (HERMES/Pgp-1) in cell trafficking, based on its
interaction with vascular endothelium, has afforded it some prominence in
the facilitation of metastatic spread. CD44 occurs in several isoforms with a
varying number of exons between the transmembrane and common N-
terminus. Normal epithelium expresses the CD44H isoform with hyaluran-
binding domains, but lacking the intervening v1—v10 exons; expression of
certain of these exons on tumors is indicative of a growth advantage, as
they are present with higher frequency on more advanced cancers. Stable
transfection of a nonmetastatic tumor with a CD44 cDNA clone
encompassing exons v6 and v7 induced the ability to form metastatic
tumors—a most striking effect. Further, injection of a monoclonal anti-
CD44 v6 prevented the formation of lymph node metastases. Exons v6 and
v10 have now been shown to bind blood group H and chondroitin 4-sulfate,
respectively, and the latest hypothesis is that these carbohydrates can bind
to CD44H on endothelium and thence homotypically to each other so
generating a metastatic nidus.
Changes have quite frequently been observed in the expression of class I
MHC molecules. For example, oncogenic transformation of cells infected
with adenovirus 12 is associated with highly reduced class I as a
consequence of very low levels of TAP- 1 and TAP-2 mRNA. Mutation
frequently leads to diminished or absent class I expression linked in most
cases to increased metastatic potential, presumably reflecting decreased
vulnerability to T-cells but not NK cells. In breast cancer, for example,
around 60% of metastatic tumors lack class I.
Spontaneous immune responses to tumors
Immune surveillance against strongly
immunogenic tumors
When present, many of the antigens discussed in the previous section can
provoke immune responses in experimental animals that lead to resistance
against tumor growth, but they vary tremendously in their efficiency.
Powerful antigens associated with tumors induced by oncogenic viruses or
ultraviolet light generate strong resistance, while the transplantation
antigens on chemically induced tumors (Milestone 17.1) are weaker and
somewhat variable; disappointingly, tumors that arise spontaneously in
animals produce little or no response. The immune surveillance theory
would predict that there should be more tumors in individuals whose
adaptive immune systems are suppressed. This undoubtedly seems to be the
case for strongly immunogenic tumors. There is a considerable increase in
skin cancer in immunosuppressed patients living in high sunshine regions
north of Brisbane and, in general, transplant recipients on
immunosuppressive drugs are unduly susceptible to skin cancers, largely
associated with papilloma virus, and EBV- positive lymphomas. The EBV-
related Burkitt’s lymphomas crop up with undue frequency in regions
infested with malarial infection, known to compromise the efficacy of the
immune system. Likewise, the lymphomas that arise in children with T-cell
deficiency linked to Wiskott–Aldrich syndrome or ataxia telangiectasia
express EBV genes; they show unusually restricted expression of EBV
latent proteins that are the major potential target epitopes for immune
recognition, while cellular adhesion molecules, such as intercellular
adhesion molecule-1 (ICAM-1) and lymphocyte function-associated
molecule-3 (LFA- 3), which mediate conjugate formation with Tc cells,
cannot be detected on their surface (Figure 17.10). Knowing that most
normal individuals have highly efficient EBV-specific Tc cells, this must be
telling us that only by downregulating appropriate surface molecules can
the lymphoma cells escape even the limited T-cell surveillance operating in
these patients.
Figure 17.10. Tumor escape mechanisms.
These examples aside, it must be acknowledged that the incidence of
spontaneous tumor formation in mice lacking T- and B-lymphocytes is not
substantially higher than in those with intact immune systems; the same is
also true for immuno-deficient humans. Such observations weaken
arguments that adaptive immune responses have a significant role to play in
cancer prevention. Nonetheless, while normal adaptive immune responses
may be insufficient to deal with the establishment of many tumors, this does
not necessarily mean that the immune system cannot be manipulated to
deliver an effective response.
A role for innate immunity?
There is an uncommon flurry of serious interest in natural killer (NK)
cells. It is generally accepted that they subserve a function as the earliest
cellular effector mechanism against dissemination of blood-borne
metastases. Let’s look at the evidence. Patients with advanced metastatic
disease often have abnormal NK activity and low levels appear to predict
subsequent metastases. In experimental animals, removal of NK cells from
mice with surgically resected B16 melanoma resulted in uncontrolled
metastatic disease and death. Acute ethanol intoxication in rats boosted the
number of metastases from an NK-sensitive tumor 10-fold, but had no
effect on an NK- resistant cancer, hinting at a possible underlying cause for
the association between alcoholism, infectious disease and malignancies.
(Those who enjoy the odd bacchanalian splurge should not be too upset—
be comforted by the beneficial effect in heart disease, but no excesses
please!) Powerful evidence implicating these cells in protection against
cancer is provided by beige mice that congenitally lack NK cells. They die
with spontaneous tumors earlier than their nondeficient +/bg littermates,
and the incidence of radiation-induced leukemia is reduced by prior
injection of cloned isogeneic NK cells that could be suppressing
preleukemic cells. Note, however, that tumors induced chemically or with
murine leukemia virus were handled normally.
Resting NK cells are spontaneously cytolytic for certain, but by no means
all, tumor targets; cells activated by IL-2 and possibly by IL-12 and IL-18
display a wider lethality. As described earlier in Chapter 4, recognition of
the surface structures on the target cell involves various activating and
inhibitory receptors, but it is important to re-emphasize that recognition of
class I imparts a negative inactivating signal to the NK cell. Conversely,
this implies that downregulation of MHC class I, which tumors employ as a
strategy to escape Tc cells (Figure 17.10-, would make them more
susceptible to NK attack. The tumor cells can fight back by expressing
CD99, which downregulates NK CD16, and the growth inhibitor RCAS1
(receptor-binding cancer antigen expressed on SiSo cells), which induces
apoptosis in NK as well as in Tc cells (Figure 17.10). It is not clear whether
surface FasL, which can repel attack by the Fas-positive cytolytic T-cells, is
also effective against NK cells, but the relative resistance of tumor cells to
apoptosis must be innately protective.
Divisions are surfacing in the NK ranks. The NK cells, which remarkably
constitute up to 50% of the liver-associated lymphocytes in humans, have a
higher level of expression of IL-2 receptor and adhesion molecules such as
integrins compared with NK cells in peripheral blood. They are precursors
of a subset of activated adherent NK cells (A-NK) that adhere rapidly to
solid surfaces under the influence of IL-2 and are distinguished from their
nonadherent counterparts by their superiority in entering solid tumors and
in prolonging survival following adoptive transfer with IL-2 into animal
models of tumor growth or metastasis. The nonadherent NK variety is
better at killing antibody-coated cancer cells through antibody-dependent
cellular cytotoxicity (ADCC), mediated by their CD16 FcγRIII receptor.
Be kind to your NK cells. Really late nights that involve major
curtailment of slow-wave sleep lead to drastic falls in NK cells and levels of
IL-2, quite apart from bleary eyes.
Tumor escape mechanisms
Strong supporting evidence for a role for the immune system in surveillance
against transformed cells comes from observations that tumors employ a
range of strategies to evade and manipulate the immune system. Indeed, it
could be said that tumors are positively brimming with various
immunological escape mechanisms (Figure 17.10- and thus they resemble
successful infections. We have already referred to the fact that
downregulation of HLA class I molecules to make the tumor a less
attractive target for cytolytic T-cells is a favorite ploy. This is a common
feature of breast cancer metastases, and this is true also of cervical
carcinoma where, prognostically, loss of HLA- B44 in premalignant lesions
is an indicator of tumor progression. Rather than lose expression of all class
I molecules and risk attracting the attentions of NK cells, tumors may lose
just the expression of class I alleles that are capable of presenting antigenic
peptides to T-cells.
Immunoediting weeds out immunogenic cells
As we have noted earlier in this chapter, many tumors are also not
particularly immunogenic to begin with, possibly because strongly
immunogenic tumors may be readily weeded out and fail to develop to the
point that they become clinically significant. In this way, the immune
system may exert a Darwinian selective pressure for cancer-causing
mutations that are largely immunologically silent: a process that has been
termed immunoediting. Subtle point mutations in oncogenes, such as RAS,
that have profound effects on the function of the protein products of such
genes and contribute to transformation, may completely fail to create any
new epitopes that would result in immune attack. In a similar vein,
complete loss of expression of important tumor suppressor genes, such as
P53 or RB, through nonsense mutations would also fail to create any new
epitopes.
Tolerization against tumor antigens due to the
absence of co-stimulation
Loss of tumor antigen epitopes, where they do arise, represents another
escape mechanism and mutations in an oncogenic virus itself can increase
its tumorigenic potential. Thus the frequent association of a high-risk
variant of human papilloma virus with cervical tumors in HLA-B7
individuals is attributed to the loss of a T-cell epitope that would otherwise
generate a protective B7-mediated cytolytic response. As outlined earlier,
there is also an increasing appreciation that tumors may create a
microenvironment where active tolerization of tumor-infiltrating
lymphocytes occurs through failure of DCs to express the appropriate co-
stimulatory molecules. Recall that DCs that deliver antigen in the absence
of proper co-stimulation, in the form of CD28 ligands, render T-cells
anergic (Figure 17.4). As we have discussed earlier in this chapter, but it is
a point worth reiterating, one reason why DCs in the vicinity of tumors may
fail to become activated may be due to the absence of PAMPs or DAMPS
that can upregulate co-stimulatory molecules on DCs that encounter these
signals (Figure 17.4). We have already noted that DCs do not exist in a
state of perpetual activation and are unable to provide proper co-
stimulation unless they encounter appropriate molecules that possess DC-
activating properties. In the context of infection, DC-activating molecules
(Toll-like receptor ligands) are derived from the infectious agent and are
usually structures that are shared by many pathogens but not found in the
host. However, tumors that release endogenous danger signals (DAMPs)
may fail to tolerize DCs and become subject to effective immune attack that
may result in tumor rejection or persistence through selection of immune
escape mutants (Figure 17.11). Conversely, tumors that fail to release
DAMPs may be simply regarded as self and may fail to elicit significant
immune responses.
There are also other reasons why the immune system may become
tolerant to a tumor. For example, many solid tumors secrete large amounts
of angiogenic factors such as vascular endothelial cell growth factor
(VEGF) that promote the development of the new blood vessels that tumors
need. Evidence also suggests that VEGF can suppress the maturation of
dendritic cells (DCs) and these immature or partially differentiated DCs
may tolerize to antigens that they find within the vicinity of the tumor.
Tumor counterattack mechanisms
Tumors can also decrease their vulnerability to cytotoxic T-cell attack by
expression of surface FasL (cf. p. 247) and a growth inhibitory molecule,
RCAS1, which react with T-cells bearing their corresponding receptors and
stop them in their tracks. As we have already noted earlier, tumors have also
frequently been found to secrete other immunosuppressive factors such as
TGFP and IL-10 (Figure 17.4). Such factors may help to keep burgeoning
immune responses at bay by inducing suppressor or regulatory T-cell
populations that inhibit responses to the tumor. Natural regulatory T-cells
(Tregs) that normally guard against the development of autoimmunity may
also hamper robust T-cell responses against tumors. It should also be borne
in mind that internal defects in the cell death machinery, that facilitated the
establishment of the tumor in the first instance, may also render such cells
resistant to the best efforts of cyto-toxic T-cells and NK cells to eradicate
them. The very existence of such “Houdini” mechanisms builds a case in
favor of the notion that the adaptive immune system has a significant role in
suppressing tumor growth and provides hope that this can be exploited in
the clinic.
Cancers that express neoantigens of low immunogenicity do not come
creeping out of the woodwork when patients are radically
immunosuppressed and, although T-cell responses can often be rescued
from tumor-infiltrating lymphocytes or relatively high numbers of tumor-
specific CD8 T-cells may be detected by the peptide-HLA tetramer
technique in peripheral blood, they may be functionally deficient due
perhaps to suppression by local IL-10 and TGFP. Mutation of p53 and its
overexpression are very common events in human cancer and are often
associated with the production of antibodies; but while these could prove to
have a diagnostic utility, it is most unlikely that they are of benefit to the
patient, the current view being that rell-mediated responses are crucial to
the attack against internal antigens expressed in solid tumors. Reluctantly,
one has to accept the view that, with tumors of weak immunogenicity, we
are dealing with low-key reactions that clearly play little role in curbing the
neoplastic process. That is not to say that these “weak” antigens cannot be
exploited for therapeutic purposes, as we shall soon see.
Figure 17.11. Danger signals may dictate whether dendritic cells prime
for T-cell responses or induce tolerance to tumors.
Growing tumors typically shed material from dead or dying cells and this
debris is picked up by local dendritic cells (DCs) that transport it to the
local lymph nodes for presentation to T-cells. (a) Where the tumor is
shedding molecules (“danger signals”) that are capable of activating DCs,
maturation of the DC occurs and these cells are capable of eliciting robust
immune responses from appropriate T-cells. Such tumors may then become
subject to immune attack by the activated T-cells resulting either in
rejection of the tumor, or “immunoediting” of the tumor to eliminate only
the cells presenting the antigen that initiated the response. (b) In the absence
of signals that activate DCs, resting DCs that encounter tumor-derived
material fail to become activated and any T-cells such DCs subsequently
encounter may become tolerant to tumor-derived antigens presented by
such DCs. (Based on Melief J.M. (2005) Nature 437, 41.)
Immune responses can paradoxically enhance
tumor growth
Although we tend to think of immune responses in purely destructive terms
—probably not unreasonably when one considers that much of the early
stages of an immune response is preoccupied with detecting and eliminating
nonself entities—a significant and often overlooked function of the immune
system is to restore normal tissue integrity after an infection has been
resolved by promoting wound healing. To this end, macrophages and other
innate immune cells secrete growth factors and other mediators that can
stimulate proliferation of local tissue and endothelium for the purposes of
replacing cells that were killed during the acute stages of infection. There is
now much evidence that tumors, through recruitment and “re-education” of
inflammatory cells to a more wound healing phenotype, can harness the
growth-stimulating properties of such cells to subvert the actions of the
innate immune system.
For example, TNFα production in the tumor microenvironment can
stimulate TNF receptor positive tumor cells leading to activation of the
NFκB transcription factor, which can have two major consequences. One
the one hand, NFκB can promote expression of additional cytokines, such
as IL-1 and IL-6, which can have autocrine growth-promoting effects on the
tumor. On the other hand, NFκB activation can also lead to expression of
multiple apoptosis-inhibitory molecules within the tumor that may render
such cells more difficult to kill. Either way, the tumor benefits from the
combined effects of TNFα exposure. In such situations, treatment with
neutralizing anti-TNF antibodies may be of therapeutic benefit. Similarly,
inhibitors of NFκB are also under evaluation as potential chemotherapeutic
drugs.
Infection and inflammation can enhance
tumor I nitiation, promotion and
progression
Aside from the tumor escape mechanisms detailed in the previous section,
there is now considerable evidence that infection and its downstream
consequences—the inflammatory response—can promote tumor
development as well as progression. Initial evidence for a role for infection
as a factor that can influence tumor growth came from studies that noted
that postoperative infections in cancer patients frequently led to rapid
growth of previously dormant metastases (i.e. secondary tumors) after
surgical resection of the primary tumor mass. This was subsequently
confirmed by LPS treatment of tumor-bearing mice, which showed that this
had a significant growth enhancing effect on the tumor as well promoting
the establishment of metastases. It is now well accepted that chronic
infection and inflammation are among the most important epigenetic
and environmental factors that can influence the establishment and
progression of certain tumors. For example, there is a significant
association between long-term alcohol abuse—which leads to inflammation
of the liver and pancreatic tissues—and cancers of the same organs.
Similarly, inflammatory bowel disease is associated with an increased risk
of colon cancer; chronic viral hepatitis is associated with liver cancer;
Helicobacter pylori infection is associated with gastric cancer; asbestos and
silica exposure are associated with persistent lung inflammation and lung
cancer.
TLR expression on tumors can enhance tumor
growth and survival
We have already encountered the role of Toll-like receptors and their
important role in immunity particularly in the initial stages of the immune
response (see Chapter 1). TLRs are predominantly expressed on cells of the
innate immune system, such as macrophages, dendritic cells, mast cells and
NK cells, although B-rells as well as endothelial cells also utilize these
receptors in several contexts. TLR stimulation by pathogen-derived
molecular patterns can lead to the secretion of several cytokines and
recruitment of inflammatory cells such as neutrophils and additional
macrophages from monocyte precursors. However, recent evidence also
suggests that tumors may harness the positive effects of TLR stimulation to
promote growth and survival. Several tumor cell types have been found to
express TLRs and evidence is accumulating to suggest that stimulation of
these receptors on tumor cells can promote tumor proliferation and survival
via autocrine signaling. This is due to TLR engagement being a potent
NFκB-activation stimulus (see Figure 1.7). NFκB activation can help the
tumor in at least two ways. First, by upregulating the expression of anti-
apoptotic gene products, such as members of the Bcl-2 family, this can
make the tumor more resilient to the oxygen and nutrient deprivation that
frequently occur in the tumor environment. Second, by promoting the
production of cytokines such as IL-1 and IL-6 that can have mitogenic
effects on certain cell types, tumors may produce their own growth-
promoting factors. The presence of microbial factors, as a source of TLR
ligands, can therefore have positive benefits for TLR-expressing tumors,
which is just one way in which persistent infection can drive tumor
progression.
An inflammatory environment can foster
mutation
Chronic infection can also influence tumorigenesis in additional ways, with
cytokine and reactive oxygen production by inflammatory cells recruited to
the site of infection also capable of promoting tumor initiation and
progression. Inflammatory cells, especially activated macrophages, can
cause DNA damage through the production of reactive oxygen and nitrogen
species and thus generate mutations that can lead to cellular transformation
(Figure 17.12). Should this occur on an occasional basis, it is tolerable and
can be viewed as one of the downsides of having a robust immune system.
Moreover, cells with DNA damage can be dealt with through DNA repair,
elimination via apoptosis, or one of the other cell-intrinsic mechanisms of
tumor suppression, as we discussed earlier in this chapter (Figure 17.3).
However, if an inflammatory response is allowed to smolder for months or
years on end, as happens in chronic colitis and viral hepatitis for example,
the inflammatory response can greatly increase the risk of malignant
transformation through the generation of genetic instability at the site of
inflammation (Figure 17.12).
Figure 17.12 Chronic inflammation can promote malignant
transformation.
Persistent, smoldering inflammation can lead to genetic instability through
recruitment of macrophages and other innate immune cells that are capable
of provoking DNA damage through the production of reactive oxygen and
nitrogen species. Persistent DNA damage can result in the generation of
mutations that can lead to cellular transformation.
Certain oncogenic mutations can drive the
production of proinflammatory cytokines and
chemokines
A substantial proportion of tumors carry mutations in Ras or its downstream
target B-Raf that render these proteins constitutively active. Constitutively
active Ras or B-Raf lead to activation of the MEK and ERK kinases
downstream, which has the effect of activating a battery of transcription
factors that can promote cell division. Among the targets of these
transcription factors are the genes for IL-6 and IL-8 and, as a consequence,
tumors carrying gain-of-function mutations in Ras and B-Raf frequently
express these cytokines. If such an event were detrimental to the tumor, one
would expect that clonal variants would emerge where the expression of IL-
6 and IL-8 was silenced. However, it appears instead that secretion of such
cytokines can enhances tumor growth in a number of ways, as we have
already alluded to earlier. One possibility is that IL-6 could have autocrine
growth and survival-promoting effects on the tumor itself, acting to enhance
cell division or lead to the expression of anti-apoptotic proteins (Figure
17.13). Another is that IL-6 acts in a paracrine fashion on surrounding
stromal cells to promote angiogenesis, thereby enhancing blood supply to
the tumor. Indeed, evidence for the latter scenario has been found in mouse
models where growth of chemical-mduced skin tumors was impaired in IL-
6 knockout mice and this was related to the effects of IL-6 on nearby
endothelial cells, rather than on the tumor itself. Use of IL-6 knockout mice
has also provided clear evidence that these animals are resistant to the
development of malignant myeloma, a malignancy affecting the B-cell
lineage. Furthermore, certain IL-6 promoter polymorphisms, which result in
the production of higher levels of this cytokine, have been found to
correlate with a poorer prognosis in breast cancer.
Similarly, tumor-derived IL-8 has been found to promote infiltration of
tumors by neutrophils and macrophages that, as discussed earlier, can
promote tumor growth through the production of other proinflammatory
cytokines such as IL-1 and TNF as well as by secreting matrix
metalloproteases that can remodel the extracellular matrix and promote
tumor spread (Figure 17.13). Activity of inflammatory cells also led to
increased recruitment of endothelial cells and promoted angiogenesis. Use
of neutralizing anti-IL-8 antibodies in Rasi nduced tumor models led to a
marked reduction in tumor growth with greatly increased tumor necrosis
apparent.
Although the fact that tumors can deliberately recruit cells of the innate
immune system by secreting chemokines and pro-inflammatory cytokines is
pretty ominous, it does suggest that one way of attacking such tumors may
be through neutralizing such factors with appropriate monoclonal
antibodies, as such antibodies are now available and have been approved
for use in other conditions such as psoriasis.
Approaches to cancer immunotherapy
Although immune surveillance seems to operate only against strongly
immunogenic tumors, the identification of a range of tumor antigens is a
positive step forward (Table 17.1), and has set the stage for exploring how
these antigens may be exploited to harness the patient’s own immune
system in the fight against cancer. On one point all are agreed: if
immunotherapy is to succeed, it is essential that the tumor load should first
be reduced by surgery, irradiation or chemotherapy, as not only is it
unreasonable to expect the immune system to cope with a large tumor mass,
but considerable amounts of antigen released by shedding would tend to
prevent the generation of any significant response in some cases due to the
stimulation of regulatory T-cells. This leaves the small secondary deposits
as the proper target for immunotherapy.
So what type of immune response is required for tumor destruction?
Studies in mouse models, as well as cancer patients, over the past decade or
so suggest that a number of criteria need to be fulfilled in order to obtain
killing of tumor cells in sufficient numbers to positively impact on the
course of disease. First, sufficient numbers of T-cells with highly avid
recognition of tumor antigens must be generated. Then, these cells must be
able to traffic to the site of the tumor and invade the stroma (supporting
cells) associated with the tumor. Finally, these lymphocytes should become
activated at the site of the tumor and be capable of engaging the tumor with
cytotoxic granules or cytokines such as TNF. Experience to date suggests
that fulfilling all of these criteria poses an immense challenge and that
immunotherapy is unlikely to offer any “magic bullet” cures. More
realistically, immunological manipulations, in tandem with conventional
chemo- and radiotherapy, are likely to be the way forward.
Figure 17.13 Activating Ras mutations can lead to proinflammatory
effects.
Oncogenic Ras and B-Raf mutations may lead to the production of
proinflammatory cytokines, such as IL-6 and IL-8, that can have diverse
pro-survival and growth-promoting effects on tumors as shown.
Table 17.1. Potential tumor antigens for immunotherapy. (Reproduced
with permission from Fong L. & Engleman E.G. (2000) Dendritic cells
in cancer immunotherapy. Annual Review of Immunology 18, 245.)
Antigen Malignancy
Tumor specific
Immunogiobuiin V-region B-cell non-Hodgkin’s lymphoma, multiple myeloma
TCR V-region T-cell non-Hodgkin’s lymphoma
Mutant p21/ras Pancreatic, colon, lung cancer
Mutant p53 Colorectal, lung, bladder, head and neck cancer
Deveopmental
p210/bcr-abi fusion Chronic myelogenous leukemia, acute lymphoblastic leukemia
product
MART-1/Meian A Melanoma
MAGE-1, MAGE-3 Melanoma, colorectal, lung, gastric cancers
GAGE famiiy Melanoma
Teiomerase Various
Viral
Human papilloma virus Cervical, penile cancer
Epstein–Barr virus Burkitt’s lymphoma, nasopharyngeal carcinoma, post-transplant
lymphoproliferative disorders
Tissue specific
Tyrosinase Melanoma
gp100 Melanoma
Prostatic acid phosphatase Prostate cancer
Prostate-specific antigen Prostate cancer
Prostate-specific Prostate cancer
membrane antigen
Thyroglobulin Thyroid cancer
α-Fetoprotein Liver cancer
Overexpressed
HER2 Breast and lung cancers
Carcinoembryonic Colorectal, lung, breast cancer
antigen Muc-1 Colorectal, pancreatic, ovarian, lung cancer
Antigen-independent cytokine therapy
The first clear indication that manipulation of the immune system could be
beneficial came from studies that utilized antigen-independent strategies to
nonspecifically boost the immune response to the tumor. Cytokines such as
IL-2, IFN and TNF have pleiotrophic effects on the immune system and
some of these have shown promise in animal models as well as in clinical
settings. Systemic toxicity has limited the utility of TNF that exhibits rapid
and severe hepatotoxicity in animal models and is therefore of limited use
in cancer therapy.
Interleukin treatment
High doses of IL-2 have been administered to patients with metastatic
melanoma or kidney cancer, and at least partial tumor regression was
observed in 15–20% of patients, with some patients displaying complete
regression. The beneficial effects of high doses of IL-2 may be due to
stimulation of preexisting tumor-responsive T-cells or due to NK activation.
On activation by IL-2 or IL-12, NK cells are capable of killing a variety of
fresh tumor cells in vitro and, on the basis of studies on mice with
mammary glands carrying the HER2/neu oncogene, it would not be
unreasonable to conduct a trial of systematic IL-12 treatment in cancer
patients with minimum residual disease in an attempt to prevent recurrence
and to inhibit incipient metastases. Because of the promising results seen
upon IL-2 administration, many subsequent tumor vaccine trials have been
conducted in combination with this cytokine.
Interferon therapy
In trials using IFNα and IFNβ, a 10–15% objective response rate was seen
in patients with renal carcinoma, melanoma and myeloma, an approximate
20% response rate among patients with Kaposi’s sarcoma, about 40%
positive responders in patients with various lymphomas and a remarkable
response rate of 80–90% among patients with hairy cell leukemia and
mycosis fungoides.
With regard to the mechanisms of the antitumor effects, in certain tumors
IFNs may serve primarily as antiproliferative agents; in others, the
activation of NK cells and macrophages may be important, while
augmenting the expression of class I MHC molecules may make the tumors
more susceptible to control by immune effector mechanisms. In some
circumstances the antiviral effect could be contributory.
For diseases like renal cell cancer and hairy cell leukemia, IFNs have
induced responses in a significantly higher proportion of patients than have
conventional therapies. However, in the wider setting, most investigators
consider that their role will be in combination therapy, e.g. with active
immunotherapy or with various chemotherapeutic agents where synergistic
action has been observed in murine tumor systems. IFNα and β synergize
with IFNγ and the latter synergizes with TNF. IFNα acts as a radiation
sensitizer and its ability to increase the expression of estrogen receptors on
cultured breast cancer cells suggests the possibility of combining IFN with
anti-estrogens in this disease.
Colony-stimulating factors
Normal cell development proceeds from an immature stem cell with the
capacity for unlimited self-renewal, through committed progenitors, to the
final lineage-specific differentiated cells with little or no potential for self-
renewal. Therapy aimed at inducing tumor cell differentiation is founded on
the idea that the induction of cell maturation decreases and possibly
abrogates the capacity of the malignant clone to divide. Along these lines,
GM-CSF has been shown to enhance the differentiation, decrease the self-
fenewal capacity and suppress the leukemogenicity of murine myeloid
leukemias. Recombinant human products are now undergoing trials.
It is over 100 years since the physician Coley gave his name to the
mixture of microbial products termed Coley’s toxin. This concoction
certainly livens up the innate immune system and does produce remission in
a minority of patients. The suggestion has been made that these beneficial
effects are due to the release of TNF as the vascular endothelium of tumors
is unduly susceptible to damage by this cytokine and hemorrhagic necrosis
is readily induced. It is questionable whether the critical levels of TNF are
reached in the human as these would be very toxic, although one study
involving perfusion of an isolated limb with TNF, IFNγ and melphalan
provoked lesions in the tumor endothelium without affecting the normal
vasculature. Opinion is coming round to the view that the Coley
phenomenon may be linked more to boosting a preexisting weak antitumor
immunity.
Stimulation of cell-mediated immune responses
The current dogma is that T-cells rather than antibodies are capable of
savaging solid tumors, particularly those expressing processed intracellular
antigens on their surface, and, as the majority are MHC class II negative, it
looks as though we are aiming at essentially CD8 cytotoxic T-cell
responses, although CD4 T-cells can be involved in protective reactions
against tumor-associated vasculature and are required for persistence of
CD8 T-cells.
Vaccination with viral antigens
Based on the observation that certain forms of cancer (e.g. lymphoma,
cervical carcinoma, hepatocellular carcinoma) are caused by oncogenic
viruses, attempts are being made to prepare suitable vaccines against these
viruses. Viruses associated with cancer include Epstein–Barr virus (EBV),
HPV, hepatitis B virus (HBV), hepatitis C virus (HCV), human T-cell
leukemia virus-1 (HTLV-1) and Kaposi’s associated sarcoma virus (KSHV).
Vaccines against several of these viruses have been in development for
some years now but progress has been hampered by the poor
immunogenicity of many of the vaccine candidates tested. Happily, some of
these vaccines have now made it through clinical trials and we have entered
the era of vaccination against several forms of cancer.
Worldwide, chronic hepatitis B infection is responsible for 80% of all
liver cancer, a major cause of mortality. Although the first HBV vaccine
became available in 1981, this vaccine was based upon inactivated pooled
plasma from infected donors and was discontinued in 1990 due to the
development of a safer, more effective, vaccine based on a recombinant
subunit approach. The HBV vaccine contains a recombinant form of one of
the viral envelope proteins, hepatitis B surface antigen (HBsAg).
Immunization generates strong neutralizing antibodies against HbsAg and
vaccination of newborns has led to a marked decrease in rates of liver
cancer. Attempts to develop a vaccine against the related HCV virus have
met with little success thus far, despite several clinical trials in recent years.
Vaccines based upon recombinant proteins, peptides, DNA encoding viral
proteins are all at various stages of development.
A very recent example of a successful vaccination strategy against a
virus-induced cancer is represented by the recent approval of two different
prophylactic vaccines targeting HPV, the major cause of cervical carcinoma
in women. HPV is endemic in the human population, with ~50% of women
becoming HPV-positive by 24 years of age, and is responsible for the
development of the majority of cervical carcinomas in women, as well as
genital, anal and penile warts. Globally, cervical cancer is the second most
common cause of cancer in women and each year almost 50% of women
that are diagnosed with cervical cancer (~500000 worldwide) die from it.
The search for a HPV vaccine started in the late 1980s and culminated in
the approval of the first preventative HPV vaccine in 2006. An additional
preventive HPV vaccine was also approved for human use a year later.
These vaccines have proved to be highly effective against the development
of cervical cancer in women. Both vaccines are composed of recombinant
L1 protein, one of the two HPV nucleocapsid proteins, derived from the
commonest HPV genotypes: HPV type 16 and 18, which are responsible for
almost 70% of cervical cancers. The L1 nucleocapsid protein assembles
into virus-like particles, which are morphologically identical to HPV
virions but are obviously noninfectious, and produces a robust neutralizing
antibody response that provides protection from HPV infection via mucosal
and epithelial surfaces.
HPV vaccination should ideally take place before infection has occurred,
which in practice means before sexual activity has begun. Although this has
not been investigated to date, it is possible that HPV vaccines may well
prove to have some benefit in the early, pre-cancerous, dysplastic stages of
cervical cancer progression, i.e. after infection has occurred.
Figure 17.14 Immunotherapy by transfection with co-stimulatory
molecules.
The tumor can only stimulate the resting T-cell with the co-stimulatory help
of B7–1 and -2 and/or cytokines such as GM-CSF, γ-interferon and various
interleukins, IL-2, -4 and -7. CTLA-4 blockade enhances immunogenicity.
Alternatively, the T-cell can be stimulated directly by tumor antigens
presented by dendritic cells (DCs) that can themselves be activated by
cross-linking their surface CD40 with antibody (see below). Once activated,
the T-cell with upregulated accessory molecules can now attack the original
tumor lacking co-stimulators.
Work is also in progress to develop a vaccine against Epstein-Barr virus
(EBV), which is responsible for the development of Burkitt’s lymphoma as
well as nasopharyngeal carcinoma, one of the most common cancers in
China. The major site of EBV infection is the oropharyngeal cavity with
transmission occurring via oral contact, hence the name “kissing disease.”
The major EBV surface glycoprotein gp350/220 is the main target of EBV
neutralizing antibodies and several vaccine candidates based on gp350/220
have been developed but generally need strong adjuvants to elicit decent
immunity. Phase II clinical trials of one of these EBV candidate vaccines
are under way.
Immunization with whole tumor cells
A variety of approaches utilizing both autologous and allogeneic whole
tumor cell preparations have been tried in an effort to awaken antitumor
responses. This has the advantage that we do not necessarily have to know
the identity of the antigen concerned. The disadvantage is that the majority
of tumors are weakly immunogenic, and do not present antigen effectively
and so cannot overcome the barrier to activation of resting T-cells.
Remember, the surface MHC-peptide complex on its own is not enough;
co-stimulation with molecules such as B7.1 and B7.2 and possibly certain
cytokines is required to push the G0 T-cell into active proliferation and
differentiation. Once we get to this stage, however, the activated T-cell no
longer requires the accessory co-stimulation to react with its target, for
which it has a greatly increased avidity due to upregulation of accessory
binding molecules such as CD2 and LFA-1 (cf. p. 215; Figure 17.14).
Whole cell immunization approaches have been largely unsuccessful in
human clinical trials, possibly because of the very limited quantity of
antigenic molecules present in whole cells where the majority of proteins
present are nonimmunogenic.
When proper co-stimulation is provided encouraging results have been
reported, at least in animal models. Vaccination with B7-transfected murine
+
melanoma generated CD8 cytolytic effectors that protected against
subsequent tumor challenge; in other words, transfection enabled the
melanoma cells to present their own antigens efficiently, while the
untransfected cells were vulnerable targets for the cytotoxic T-cells so
produced. A further telling observation was that an irradiated
nonimmunogenic melanoma line that had been transfected with a retroviral
vector carrying the GM-CSF gene stimulated potent and specific antitumor
immunity, almost certainly by enhancing the differentiation and activation
of host antigen-presenting cells.
A less sophisticated but more convenient approach ultimately utilizing
similar mechanisms involves the administration of the irradiated melanoma
cells together with BCG that, by generating a plethora of inflammatory
cytokines, increases the efficiency of presentation of tumor antigens derived
from necrotic cells. In a large-scale study of over 1500 patients, 26% of
vaccinees were alive at 5 years compared with only 6% of those treated
with the best available conventional therapy. It would be exciting to
suppose that in the future we might expose a tumor surgically and then
transfect it in situ by firing gold particles (cf. p. 179) bearing appropriate
gene constructs such as B7, IFNγ (to upregulate MHC class I and II), GM-
CSF, IL-2, and so on (Figure 17.14). There is a real risk of inducing
autoimmune responses to cryptic epitopes shared with other normal tissues
that the prudent investigator will not overlook.
Therapy with subunit vaccines
The variety of potential tumor antigens thus far identified (Table 17.1) has
spawned a considerable investment in clinical therapeutic trials using
peptides as vaccines. Because of the pioneering work in characterizing
melanoma-specific antigens, this tumor has been the focus of numerous
studies that exploit to the full the academic background to modern
immunology. Encouraging results in terms of clinical benefit, linked to the
generation of cytolytic T-cells (CTLs), have been obtained following
vaccination with peptides complexed with heat-shock proteins or modified
at class I anchor residues to improve MHC binding. Such peptides have
been delivered either alone, using recombinant viruses (fowlpox,
adenovirus, vaccinia), or as naked DNA, along with adjuvant. The inclusion
of accessory factors, such as IL-2 or GM-CSF, and CTLA-4 blockade can
be crucial for success. Potentially tolerogenic peptide vaccines can be
converted into strong primers for CTL responses by triggering CD40 with a
cross-Unking antibody that can substitute for T-cell help in the direct
activation of CTLs (Figure 17.15). Anti-CD40 treatment alone was also
shown to partially protect mice bearing CD40-negative lymphoma cells, an
effect attributed to the activation of endogenous dendritic antigen-
presenting cells (cf. Figure 17.14). However, although some promising
indications of immune responses to tumors have been recorded using such
approaches, a recent evaluation of multiple vaccine-based clinical trials
involving 440 patients, mainly suffering from melanoma, produced an
objective response rate of only 2.6%. This disappointingly poor statistic is
rather sobering and suggests that we still have some way to go before
optimism is warranted. It would be premature to write – off vaccination
approaches at this stage, however, as it should be borne in mind that all of
the clinical trials that have been carried out using such vaccines have been
conducted in patients with advanced disease. Moreover, all standard
therapies have, more often than not, also failed in such individuals.
Vaccination with tumor antigens may prove to be more successful where
early diagnosis has occurred, or where a genetic predisposition towards a
familial form of cancer exists, as a preventative measure against tumor
development.
Figure 17.15 CD40 ligation enhances the protective effect of a peptide
vaccine against a pre-existing tumor.
Six days after injection of human papilloma virus-16 (HPV16)-transformed
syngeneic cells, mice were immunized with the HPV16-E7 peptide in
incomplete Freund’s adjuvant with or without an anti-CD40 monoclonal, or
left untreated. (Data from Diehl L. et al. (1999) Nature Medicine 5, 774,
reproduced with permission.)
Enhancing immunoreactivity of existing tumor
infiltrating lymphocytes
As mentioned earlier, tumors frequently exhibit the presence of infiltrating
lymphocytes that appear anergic. One means of reactivating such cells may
be to neutralize signals that serve to dampen TCR triggering, CTLA-4
being a prime example. Recall from Chapter 8 that CTLA-4 can inhibit
TCR co-stimulation by raising the threshold for successful TCR activation,
as a result of competing for CD80/CD86 molecules on the DC. Thus,
blockade of CTLA-4 interaction with its ligands may boost the capacity of
tumor infiltrating lymphocytes to get on with the job of attacking the tumor
and this is currently under exploration in cancer patients.
A related strategy involves inhibiting members of the c-Cbl ubiquitin
ligase family that have been implicated in desensitizing the T-cell receptor
to peptide-MHC stimulation. Recall from Chapter 8 that Cbl-b-deficient
naive T-cells do not have a requirement for CD28-dependent co--timulation
for productive activation. This appears to be due to the role of Cbl-b in
suppressing certain TCR-initiated signals under normal circumstances,
thereby raising the threshold for T-cell activation. Thus, strategies aimed at
selectively inhibiting Cbl-b in tumor-infiltrating T-cells may sufficiently
lower their threshold for activation, such that productive immune responses
now ensue without recourse to co-stimulation.
Figure 17.16 Improving the efficacy of adoptive cell transfer-based
immunotherapy.
A variety of strategies are being employed to enhance the efficacy of
adoptive therapy using ex vivo expanded T-cells. (a) Tumor-reactive T-cells
(dark blue) from the patient are stimulated in vitro with APCs. To enhance
stimulation of tumor-reactive T-cells antigen-presenting cells (APCs) can be
transfected with genes encoding tumor antigens. (b) Selection of tumor-
(
reactive T-cell clones or lines can be enhanced using peptide-MHC
tetramers or bispecific antibodies to stimulate specific T-cell precursors.
(c,d) Tumor-specific cells are then expanded in IL-2 followed by (e)
intravenous infusion of tumor-specific T-cells into the patient. (f)
Successful persistence of the transferred T-cells may be enhanced by prior
depletion of host lymphocytes and/or administration of homeostatic
cytokines (IL-2, IL-15, IL-21) post-infusion. (Based upon Riddell S.R.
(2004) Journal of Experimental Medicine 200, 1533.)
Adoptive T-cell transfer
Adoptive cell transfer (ACT) with large numbers of ex vivo expanded T-
cells may overcome some of the barriers to effective therapy seen with
conventional vaccination approaches (Figure 17.16). It may even be
possible to genetically engineer the adoptively transferred cells to
constitutively express cytokines such as IL-2 or GM-CSF to boost their
activity. The generation of cytotoxic T-cell effectors ex vivo has the
potential to uncover responses that are not evident in an environment where
tumor-derived inhibitory factors, or T-regulatory cells, may be present. The
typical approach involves isolating T-cells from patients and these are then
expanded in vitro in the presence of high concentrations of IL-2 (Figure
17.16). To maximize the chances of expanding rare tumor-reactive T-cell
precursors, mature DCs expressing co-stimulatory signals along with a
source of tumor antigen are now in common use. Over a period of 2–3
weeks 1000-fold expansion of T-cells can be achieved. These in vitro
expanded CD8 T-cells are then transferred back to the patient (up to 10 11
cells per individual!) but can rapidly disappear if the tumor burden is high.
Administration of IL-2 in vivo or co-transfer of CD4 T-cells can improve
CD8 T-cell survival; the presence of CD4 T-cells appears to be crucial
for persistence of CD8 T-cells and optimal cytotoxic effector function. The
failure of vaccination approaches using predominantly class I-based
peptides may be due to the lack of CD4 T-cell expansion and this should be
perhaps borne in mind for future studies. ACT of in vitro expanded
lymphocytes into lymphodepleted hosts can result in up to 75% of
circulating T-cells with antitumor activity, way beyond the levels seen with
peptide vaccines. Although the numbers of individuals that have received
such ACT-based therapy are still low, very impressive objective response
rates of 40–50% have been reported in lymphodepleted melanoma patients,
with persistence of transferred cells for up to 4 months. Clearly some risks
must be borne in mind when transferring such large numbers of activated
lymphocytes into a patient, not least the possibility of generating
autoimmunity to tissues other than the tumor. Careful selection of tumor
antigens to favor those that are not expressed, or are minimally expressed,
on tissues other than the tumor is clearly essential in these situations.
There are some indications that lymphocyte-mediated tumor eradication
may be simply a numbers game. While peptide vaccination approaches can
increase circulating tumorreactive cells five- to ten-fold, this pales in
comparison with observations that up to 40% of circulating CD8 T-cells are
reactive against EBV in patients with infectious mononucleosis. Early
indications suggest that ACT is capable of achieving such impressive
numbers of specific T-cells, especially when combined with prior
lymphodepletion. The lymphopenic environment may be favorable as this
may free-up space in the lymphoid compartment for the incoming T-cells
and create less competition for homeostatic cytokines such as IL-7 and IL-
15. Another advantage of this approach is that depletion of recipient
lymphocytes can remove suppressor/regulatory T-cells that are suspected to
play a significant part in damping down antitumor responses in the first
place.
NK cell therapy
We have already alluded to the possible importance of NK cells in tumor
surveillance and tumor killing, so it is natural to consider that in vivo
expansion or adoptive transfer of large numbers of activated NKs may also
be of clinical benefit. NK-based therapies are somewhat lagging behind T-
cell-based approaches although they are not being overlooked. Clinical
trials on cancer patients have assessed the effects of daily subcutaneous
administration of low-dose IL-2, following highdose cytotoxic
chemotherapy, for its effects on NK cell numbers and activation status in
these individuals. While NK cell expansion was seen, these cells did not
appear to be maximally cytotoxic, perhaps because of inhibitory NK
receptors finding the appropriate ligands on the tumor. More recent attempts
involved using NK cells from related haploidentical donors to treat poor
prognosis patients with acute myeloblastic leukemia. The idea here is to
achieve a partial mismatch between the donor NKs and the recipient that
may provoke NK activation and greater tumor kill as a result. Expansion
and persistence of the donor NK cells was observed after high-dose
immunosuppression of recipients and complete remission in five out of 19
patients was achieved—encouraging signs indeed.
Figure 17.17 Clinical response to autologous vaccine utilizing dendritic
cells pulsed with idiotype from a B-cell lymphoma.
Computed tomography scans through patient’s chest: (a) prevaccine and (b)
10 months after completion of three vaccine treatments. The arrow in (a)
points to a paracardiac mass. All sites of disease had resolved and the
patient remained in remission 24 months after beginning treatment.
(Photography kindly supplied by Professor R. Levy from the article by Hsu
F.J. et al. (1996) Nature Medicine 2, 52; reproduced by kind permission of
Nature America Inc.)
Dendritic cell therapy
The sheer power of the dendritic cell (DC) for the initiation of T-cell
responses has been the focus of an ever-burgeoning series of
immunotherapeutic strategies that have elicited tumor-specific protective
immune responses via injection of isolated DC loaded with tumor lysates or
tumor antigens or peptides derived from them. Considerable success has
been achieved in animal models and increasingly with human patients
(Figure 17.17). The copious numbers ofDCs needed for each patient’s
individual therapy are obtained by expansion of CD34-positive precursors
in bone marrow by culture with GM-CSF, IL-4 and TNF, and sometimes
with extra goodies such as stem cell factor (SCF) and Fms-hke tyrosine
kinase 3 (Flt3) – ligand. CD14-positive monocytes from peripheral blood
are easier to access, and generate DC in the presence of GM-CSF plus IL-4;
however, they need additional maturation with TNFα that increases cost and
the chance of bacterial contamination. Another approach is to expand the
DCs in vivo by administration of Flt3-igand. The circulating blood DCs
increase in number 10–30-fold and can be harvested by leukopheresis.
Some general points may be made. First, where peptides are used to load
the DC, sequences that bind strongly to a given MHC class I haplotype
must be identified; sequences will vary between patients with different
haplotypes and they may not include potential CD4 helper epitopes.
Recombinant proteins will overcome most of these difficulties, and a
mixture should be even better as it should recruit more CTLs and be more
able to “ride out” any new tumor antigen mutations. However, proteins
taken up by DCs are relatively inefficient at “cross-priming” CD8 CTLs
through the class I processing pathway, although several tactics are being
explored to circumvent this problem: they include conjugation to an HIV-tat
“transporter” peptide that increases class I presentation 100fold and
transfection with RNA and recombinant vectors such as fowlpox. Second,
the procedure is cumbersome and costly but, if it becomes common, it will
be streamlined and, anyway, the costs must be set against the expenses of
conventional therapy and the immeasurable benefit to the patient. Third,
why does the administration of small numbers of antigenpulsed DCs induce
specific T-cell responses and tumor regression in patients in whom both the
antigen and DCs are already plentiful? As we discussed earlier in this
chapter, the suggestion has been made that DCs in or near malignant tissues
may be defective or immature, perhaps due to vascular endothelium growth
factor (VEGF) or IL-10 secretion by the tumor that may arrest DC
maturation to generate immature “tolerogenic” DCs. Such immature DCs
may smother tumor-reactive T-cell responses at birth rather than nurture
them. Along these lines, recent evidence suggests that melanoma tumors
+
+
+
secreting the CCL21 chemokine recruit CD4 CD25 Foxp3 regulatory T-
cells, shifting the tumor environment to a tolerogenic as opposed to an
immunogenic one. Alternatively, immature DCs that capture antigen in the
vicinity of a tumor, in the absence of appropriate Toll receptor ligands
(PAMPs) or danger signals (DAMPs), may simply not function as effective
antigen-presenting cells (Figure 17.4). It will certainly be interesting to see
whether direct treatment of patients with Flt3-ligand plus accessory
cytokines or intratumoral transfection with MIP-3a (CCL20) (cf. p. 235),
which attracts immature DCs, can initiate an antitumor response through
maturation and activation of endogenous DCs.
Vaccination against neovascularization
Solid tumors are composed of malignant cells as well as a variety of
nonmalignant cell types, collectively called stromal cells, such as
endothelial cells and fibroblasts. Because solid tumors cannot grow to any
appreciable size without a blood supply, tumors stimulate the production of
new blood vessels by secreting angiogenic factors, such as VEGF, that
stimulate endothelial cell proliferation. Because growing tumors are highly
reliant on their blood supply, attacking the tumor vasculature by targeting
antigens selectively expressed on these blood vessels may deprive the
tumor of oxygen and nutrients, provoking regression one would hope.
VEGF, one of a family of angiogenic factors, exerts its effects through
interaction with its cognate receptor, VEGF-R2 (also known as KDR in
humans and Flk-1 in the mouse), which provides signals that promote
proliferation, survival and motility of endothelial cells. Antibodies directed
against VEGF-R2, or indeed VEGF itself, can block tumor angiogenesis in
murine tumor models but translation into the clinic has been hampered due
to problems relating to delivery of sufficient amounts of these agents to
fully block VEGF-R2 activity. An alternative strategy involves breaking
immune tolerance to VEGF-R2-positive endothelial cells by pulsing in
vitro generated DCs with soluble VEGF-R2 followed by transferring these
cells back into the animal. A major advantage of this approach is that the
tumor endothelium, unlike the tumor itself, is genetically stable as it
represents nontrans-formed tissue, and this makes it unlikely that mutant
cells will arise that have lost VEGF-R2 expression. This strategy has been
reported to generate VEGF-R2-specific neutralizing antibody as well as
cytotoxic T-cells capable of effectively destroying endothelial cells.
Treatment of leukemia
Radical cytoablative treatment of leukemia patients using
radiochemotherapy will destroy bone marrow stem cells. These can be
removed prior to treatment, purged of any leukemic cells with cytotoxic
antibodies, and reinfused subsequently to “rescue” the patient (Figure
17.18). However, not all leukemic cells are eliminated by this treatment,
and a more effective strategy is transplantation of allogeneic bone marrow
from reasonably MHC-compatible donors that exerts an important, albeit
not completely understood, graft-versus-leukemia effect. Purging the bone
marrow ofT-cells to prevent graft-versus-host (GVH) disease, which is a
serious complication of such transplants, would at the same time remove
the prized antileukemic activity—a dilemma. “Suicide gene therapy” could
provide a solution as illustrated in Figure 17.19. Stem cells from the T-cell-
purged allogeneic bone marrow are given together with donor T-cells
transfected with herpes simplex virus thymidine kinase. The T-cells provide
factors that facilitate engraftment, defense against viral infection and the
graft-versus-leukemia action at a time when the recipient patient will have a
low tumor burden. With time, as GVH disease develops, the dividing
aggressor donor T-cells can be switched off by administration of ganciclovir
through the mechanism explained in the legend to Figure 17.19.
Figure 17.18 Treatment of leukemias by autologous bone marrow
rescue.
By using cytotoxic antibodies to a differentiation antigen present on
leukemic cells and even on other normal differentiated cells, but absent
from stem cells, it is possible to obtain a tumor-free population of the latter
that can be used to restore hematopoietic function in patients subsequently
treated radically to destroy the leukemic cells. Another angle is positive
selection of stem cells utilizing the CD34 marker.
An alternative that avoids GVH disease altogether is to inject the purged
bone marrow together with an allogeneic cytotoxic T-cell clone specific for
a leukemia-associated peptide presented by the MHC allele of the
prospective recipient patient. This usually works because the residues on
the MHC helices that contact the T-cell receptor are relatively conserved
(unlike those within the groove), so that the allo-T-cells can recognize the
MHC-peptide complex from the leukemia. Potential targets are cyclin-D1
and mdm-2 that are over-expressed in tumor cells and, in leukemic cells in
particular, the transcription factors WT-1 and GATA1 and the differentiation
antigens myeloperoxidase and CD68, which are expressed exclusively in
hematopoietic cells and are likely to have established tolerance in the
patient but not in the allogeneic CTL donor (who will have been exposed to
different processed peptides) (cf. p. 428). A rather masterful development
would be to transfect the recipient with the genes encoding the T-cell
receptor of the allo-CTL clone.
Figure 17.19 Treatment of leukemia with allogeneic bone marrow
transfer.
T-depleted allogeneic marrow provides the stem cells to “rescue” the patient
treated with cytoablative therapy. T-cells from the donor, transfected with
thymidine kinase (TK), help engraftment, provide protection against
infection and eliminate residual tumor cells by a graft-versus-leukemia
effect. The alloreactive cells eventually produce graft-versus-host disease
and can be eliminated by administration of ganciclovir. This is converted by
the TK into a nucleoside analog that ultimately becomes toxic for dividing
cells. The alternative shown is to destroy leukemic cells by supplementing
purged allogeneic marrow with a leukemia peptide-specific CTL clone
produced in third party T-cells. (Based on articles by Cohen J.L., Boyer O.
& Klatzmann D. (1999) Immunology Today 20, 172; and Stauss H.J. (1999)
Immunology Today 20, 180.)
Passive immunotherapy with monoclonal
antibodies
After many false dawns, monoclonal antibodies are finally delivering on
their early promise and some of the most promising results from
immunotherapeutic approaches to cancer treatment have been achieved
with humanized monoclonal antibodies. As detailed earlier (see Chapter 6),
early attempts to use mouse monoclonal antibodies for therapeutic purposes
were severely hampered by strong immune responses against the foreign
sequences within the mouse antibody, the so-called human anti-mouse
antibody (HAMA) response. Furthermore, mouse antibodies frequently
failed to activate desirable cytotoxic actions against the tumor, such as
complement activation and ADCC. These early difficulties have now been
overcome with the result that numerous “humanized” monoclonal
antibodies (see Chapter 6 for further details of how antibodies can be
engineered in various ways) have now entered clinical trials and 12
monoclonal antibodies targeting cell surface receptors have been approved
for therapeutic use in cancer (Table 17.2). Earlier in this chapter we
discussed the example of HER2 (human epidermal growth factor receptor
2; also called Neu or erb-B2), a member of the epidermal growth factor
receptor family, and its overexpression in a subset of breast cancers (Figure
17.6). This discovery led to the development of monoclonal antibodies
targeting HER2 (Herceptin), which proved to be effective against tumors
®
overexpressing this antigen, leading to the approval of Herceptin for
cancer therapy in 1998. Following on from this success, antibodies directed
against a variety of other cell surface molecules, such as CD20, EGF
receptor and VEGF have been approved for therapeutic use and a raft of
others are currently in the clinical pipeline. Moreover, the NIH clinical trials
website currently lists more than 1600 clinical trials, either ongoing or
completed, involving monoclonal antibodies targeting various cell surface
antigens in cancer. Thus, it is very likely that we will see the introduction of
numerous additional monoclonal antibodies to clinical practice in the
coming years, either as stand alone therapeutic agents, or as adjuncts to
conventional chemotherapy.
HER2-directed antibodies appear to work through disrupting growth-
promoting signals propagated via this EGF family receptor. Although
HER2 does not appear to bind to EGF directly (a ligand for this receptor
has yet to be identified), it is capable of forming heterodimers with other
members of the same receptor family that do bind EGF. Thus, antibodies
targeting the HER2 receptor presumably interfere with EGF-dependent
growth factor signaling, as well as spontaneous HER2 signals that are
generated as a result of its elevated surface expression, thereby provoking
growth arrest rather than death of HER2-positive tumors. However, through
use of mice lacking Fcγ receptors, a major role for ADCC responses
mediated by NK cells has also been implicated in the mode of action of
Herceptin.
Table 17.2. Selected mAbs approved or in late – stage clinical trials for
cancer therapy. (From data of Adams G.P. & Weiner L.M. (2005)
Nature Biotechnology 23 , 1147.)
In general, antibodies reacting with antigens on the surface of tumor cells
can also protect the host by complement-mediated opsonization and lysis
(modified by host complement regulatory proteins) and through recruitment
of macrophage and NK ADCC function by engagement of FcγRIII
receptors, although for macrophages this is partially countered by inhibitory
FcγRII signals. These FcR-bearing cells serve not only as cytotoxic
effectors but also as multivalent surfaces that hyper-cross-link antibody-
coated target cells so providing, in many cases, a transmembrane signal that
leads to apoptosis or premature exit from the cell cycle. This effect appears
to sensitize neoplastic cells to irradiation and DNA-damaging
chemotherapy and holds out the exciting prospect of novel synergistic
treatments whose efficacy may be enhanced by the increased
immunogenicity of the dying cells.
Immunologists have long been entranced by the idea of eliminating tumor
cells by specific antibody linked to a killer molecule and there is a truly
impressive array of ingenious initiatives. It is axiomatic that multimeric
fragments bind much more avidly than monomeric fragments due
principally to the lower off-rates (cf. p. 119)- and that constructs in the 60–
120 kDa range are optimal for targeting solid tumors—too large and
penetration is difficult, too small and kidney secretion is excessively fast.
Monovalent fragments include Fv, scFv selected by antigen from phage
libraries (cf. p. 146) and V domains based on the large CDR loops of the
H
camel and llama heavy chain antibodies. For polymers we have bivalent
and bispecific (think about the difference) diabodies (cf. p. 148) – trivalent
and trispecific triabodies, even tetrabodies, and Fabs have been linked into
dimers or trimers.
Therapeutic immunoconjugates
While antibody alone is sometimes effective, immunoconjugates are where
the most exciting developments have been made, particularly with respect
to solid tumors. Therapeutic immunoconjugates consist of a tumor-targeting
antibody linked with a toxic effector component, which can either be a
radioisotope, a toxin or a small drug molecule. Initial attempts to treat
tumors with such immunoconjugates proved disappointing, mainly because
the cytotoxic payloads conjugated to the mAbs were conventional
chemotherapeutic drugs (such as doxorubicin) that are not sufficiently toxic
when delivered in small doses. Dosimetry studies using -
adioimmunoconjugates indicate that very modest amounts, between
0.01% and 0.001% of the administered antibody per gram of solid tumor,
actually reach the tumor site. So, if the initial dose delivered to the patient is
10 micromolar, which would be a pretty high dose of most cytotoxic
compounds, and less than 0.01% of this dose is actually delivered to the
tumor, this effectively means that the effector drug or toxin has to work in
the picomolar range. The nature of the problem can be grasped when one
considers that many conventional chemotherapeutics are effective in the
micromolar or high nanomolar range.
This limitation prompted a search for much more toxic molecules to act
as conjugates and toxins seemed to fit the bill initially. Protein toxins such
as pseudomonas exotoxin and diphtheria toxin are highly toxic in vitro and
display activity in animal models, but they also proved to be highly
immunogenic in humans and rapidly induce neutralizing antibody responses
that limit their efficacy and the ability to administer repeated doses: a
problem known as the human anti-toxin antibody (HATA) response. In
some cases, practically 100% of patients developed HATA responses by
their second treatment with a toxin immunoconjugate. Quite apart from
HATA, another disadvantage of immunotoxin conjugates is a syndrome
that appears to result from nonspecific toxin-induced damage to
endothelium, called vascular leak syndrome, which also reduces the
maximum tolerated doses of such conjugates that can be used. However,
where patients are severely immunosuppressed, in the case of hematologic
malignancies for example, immunotoxin conjugates are of benefit; very
impressive complete remission rates approaching 70% have been recorded
for patients with hairy cell leukaemia using an anti-CD22—pseudomonas
toxin conjugate.
Another approach that has been pursued for several years now aims to
exploit the cytotoxic properties of radionuclides, such as iodine-131 and
yttrium-90, to irradiate the tumor in a highly precise manner. Several
clinical trials have been conducted with such radioimmunoconjugates, and
while there have been some notable successes, 90 Y- and 131 I-labeled anti-
CD20 conjugates for non-Hodgkins lymphoma for example, the results
have been generally disappointing. It has proved difficult to achieve
therapeutic efficacy with many radioimmunoconjugates without exceeding
the maximum tolerated dose, and side-effects such as myeloablation are
frequently seen. Attempts have been made to reduce these nonspecific toxic
effects by using α particle emitters, such as astatine-211, that have much
shorter path lengths than β-emitters that reduces collateral damage to other
cells. Such manipulations have the desired effect, with up to 1000-fold
higher absorbed dose ratios in target organs with α-emitters relative to their
β-emitter counterparts. But every silver lining has a cloud, or so it seems;
the α particle radioimmunoconjugates have half-lives ranging from 60
minutes to a few hours or so, making them impractical for routine clinical
use.
The search for toxic molecules in the high picomolar range eventually
paid off with the discovery of inhibitors of tubulin polymerization such as
auristatin and molecules that cause DNA double-stranded breaks such as
calicheamicin and esperamicin. One very attractive feature of these agents
is that conjugation of the drug to the antibody frequently converts it into a
pro-drug that requires removal from the antibody to regain activity.
Because the linker between drug and antibody is stable in the blood, the
conjugate exhibits virtually no toxicity until it becomes bound and
internalized by an antigen-positive target cell. Many such drug
immunoconjugates are currently in clinical trials or have been approved
for a range of cancers, including: acute myeloid leukemia (anti-CD33-
calicheamycin), colorectal and pancreatic cancer (anti-CanAg-DM1), small
cell lung carcinoma (anti-CD56-DM1) and several other malignancies (anti-
HER2/Neu-DM1). Considerable effort is also underway to develop even
more potent cytotoxic compounds for the preparation of drug
immunoconjugates. Because of their stability, potency and clinical utility,
small drug immunoconjugates are likely to rule the roost within a short
time.
Attack on the tumor blood supply
For solid tumors, the focus is upon two main targets. The first would be
minimal residual micrometastases in the bone marrow that occur in one-
third to one-half of patients with epithelial cancer after curative radical
treatment of the primary lesion. The second would be the reactive tissue
evoked by the malignant process, such as stromal fibroblasts expressing
the F19 glycoprotein and newly formed blood vessels.
As we discussed earlier, tumors generally cannot grow beyond 1 mm in
diameter without the support of blood vessels that the tumor promotes
formation of by secreting angiogenic factors such as VEGF. New blood
vessels are biochemically and structurally different from normal resting
blood vessels and so provide differential targets for therapeutic monoclonal
antibodies, even though the cancer cells themselves in a solid tumor are less
vulnerable to antibodies directed to specific antigens on their surface. Thus,
receptors for VEGF and Eph, oncofetal fibronectin, matrix metalloproteases
MMP-2 and MMP-9 and the pericyte markers aminopeptidase A and the
NG2 proteoglycan are all highly and selectively expressed in vasculature
undergoing angiogenesis. Consequently, considerable effort has been
expended in the direction of angiogenesis inhibitors such as humanized
monoclonal antibodies against VEGF and its main receptor VEGF-R2.
A noteworthy maneuver, which is unexpectedly successful, is to identify
peptides that home specifically to the endothelial cells of certain tumors by
injecting peptide phage libraries in vivo. One of the panel of peptide motifs
that has emerged from this probing strategy includes RGD in the cyclic
peptide CDCRGDCFC, a selective binder of the α β - and α β -integrins
V 3
V 5
known to be upregulated in angiogenic tumor endothelial cells. For
therapeutic exploitation, these peptides can be linked to appropriate drugs,
such as doxorubicin, or a pro-apoptotic peptide. Overall, there are
undoubtedly a substantial number of targets for the “magic bullets.”
Inhibition of the production of proinflammatory
cytokines in the tumor environment
Based on the observation that production of proinflammatory cytokines by
tumor-associated macrophages and fibroblasts can frequently be beneficial
for the tumor, there may be instances in which neutralizing antibodies
towards IL-6 and TNF, as well as other proinflammatory cytokines may
have beneficial effects in terms of reducing the blood supply and the
stromal support network in the vicinity of the tumor. Studies in mice have
shown that neutralizing antibodies against TNF, as well as NFκB inhibitors
can have protective effects in colon and breast cancer models.
SUMMARY
Cellular transformation
Cancer is typically caused by genetic lesions that affect genes that
promote proliferation in tandem with lesions that interfere with the
elimination of cells through apoptosis.
Cancer is not a single disease and represents a wide spectrum of
conditions caused by a failure of the controls that normally govern cell
proliferation, differentiation and cell survival.
Cellular transformation is a multi-step process and involves the
acquisition of a series of mutations in oncogenes and tumor suppressor
genes that cooperate to achieve the fully transformed state.
Cancer incidence varies between tissues.
Mutagenic agents, including viruses, promote cellular
transformation.
A variety of cell-intrinsic mechanisms of tumor suppression exist
The requirement for growth factors normally prevents uncontrolled
growth.
Telomere shortening acts as a barrier to cellular transformation.
Tumor suppressor proteins monitor cell division and can deploy a
range of countermeasures upon detection of DNA damage or aberrant
mitogenic signaling, including: DNA repair, premature cellular
senescence or apoptosis.
The cancer problem from an immune perspective
Transformed cells are not usually highly immunogenic are therefore
not recognized by cells of the immune system.
Cancers lack PAMPs and contain few nonself determinants.
Lack of T-cell co-stimulation can tolerize to tumor antigens.
Inflammatory cytokines can enhance tumor growth.
Tumor antigens
Many candidate tumor antigens have now been identified but most
are specific to an individual tumor and are not shared between
individuals.
Processed peptides derived from oncogenic viruses are powerful
MHC-associated transplantation antigens.
Some tumors express genes that are silent in normal tissues:
sometimes they have been expressed previously in embryonic life
(oncofetal antigens).
Many tumors express weak antigens associated with point mutations
in oncogenes such as ras and p53. Peptides presented by heat-shock
proteins 70 and 90 represent the unique chemically induced tumor
antigens. The surface Ig on chronic lymphocytic leukemia (CLL) cells
is a unique tumor-specific antigen.
Dysregulation of tumor cells frequently causes structural
abnormalities in surface carbohydrate structures.
The v6 and v10 exons of CD44 are intimately involved with
metastatic potential. Loss of blood group A determinants leads to a
poor prognosis.
Immune response to tumors
T-cells generally mount effective surveillance against tumors
associated with oncogenic viruses or UV induction that are strongly
immunogenic.
More weakly immunogenic tumors are not controlled by T-cell
surveillance, although sometimes low-grade responses are evoked.
NK cells probably play a role in containing tumor growth and
metastases. They can attack MHC class I-negative tumor cells because
the class I molecule imparts a negative inactivation signal to NK cells.
The A-NK subset, which expresses high levels of adhesion molecules,
is more cytolytic for fresh tumor cells.
Tumors utilize a variety of mechanisms to escape host immune
responses that suggests that the immune system exerts selective
pressure on tumors.
Infection and inflammation can enhance tumor initiation,
promotion and progression
TLR expression on tumors can enhance tumor growth and survival
through harnessing NFκB-dependent upregulation of anti-apoptotic
proteins that can make tumors resistant to stress. TLR-driven NFκB
activation can also lead to the production of cytokines, such as IL-1
and IL-6, which can have autocrine growth-promoting effects.
An inflammatory environment can foster mutation through the
production of reactive oxygen and nitrogen species that can provoke
DNA damage and generate mutations that drive cellular
transformation.
Certain oncogenic mutations can promote the production of pro-
inflammatory cytokines and chemokines, thereby recruiting cells of
the innate immune system that can enhance tumor proliferation, the
growth of new blood vessels (angiogenesis) and tumor spread.
Approaches to cancer immunotherapy
Immunotherapy is only likely to work after a tumor mass has been
debulked.
Innate immune mechanisms can be harnessed. High concentrations
of IL-2 can enhance responses to malignant melanoma and other
tumors, systemic IL-12 may be effective against minimal residual
disease. IFNy and IFNP are very effective in the T-cell disorders, hairy
cell leukemia and mycosis fungoides, less so but still significant in
Kaposi’s sarcoma and various lymphomas; they may be used in
synergy with other therapies. GM-CSF enhances proliferation and
decreases leukemogenicity of murine myeloid leukemias.
Cancer vaccines based on oncogenic viral proteins are likely to be
effective and will provide a prophylactic measure against virus-
induced cancers, such as cervical cancer.
Weakly immunogenic tumors provoke anticancer responses if given
with an adjuvant, such as BCG, or if transfected with co-stimulatory
molecules, such as B7 and cytokines IFNy, IL-2, -4 and -7.
CD8 CTLs are favored for the attack on solid tumors, and CD4 T
helper cells are likely to be required for persistence and optimal
effector function of CD8 T-cells.
A variety of potential tumor antigens have been identified and
intense effort is being expended in the investigation of peptides as
subunit vaccines. Their immunogenicity can be enhanced by
complexing with heat-shock proteins and by accessory factors such as
GM-CSF, CTLA-4 blockade and anti-CD40 stimulation.
Clinical trials using peptide-based vaccines have been disappointing
but adoptive cell transfer-based immunotherapy using in vitro
expanded CD8 T-cells has shown more promise.
Powerful immunogens have been created by pulsing dendritic
antigen-presenting cells with peptides from melanoma antigens and
framework regions of CLL Ig.
A graft-versus-leukemia effect is achieved by allogeneic CTLs or by
allogeneic bone marrow transplantation with measures to limit graft-
versus-host disease.
Monoclonal antibodies conjugated to drugs, toxins or radiolabels can
target tumor cells or antigens on new blood vessels or the reactive
stromal fibroblasts associated with malignancy. Impressive,
therapeutic results have been obtained with antibodies to CD20 in B-
cell lymphoma, CD33 in myeloid leukemia, anti-MUC-1 in ovarian
cancer and c-erbB2 overexpressed on breast cancers. Bifunctional
antibodies can bring effectors such as NK and Tc close to the tumor
target.
FURTHER READING
Ancrile B.B., O’Hayer KM. & Counter C.M. (2008) Oncogenic Ras-
induced expression of cytokines: a new target of anti-cancer therapeutics.
Molecular Interventions 8, 22–27.
Banchereau J. & Palucka A.K. (2005) Dendritic cells as therapeutic
vaccines against cancer. Nature Reviews Immunology 5, 296–306.
Begent R.H.J. et al. (1996) Clinical evidence of efficient tumor targeting
based on single-chain Fv antibody selected from a combinatorial library.
Nature Medicine 2, 979–984.
Chen R., Alvero A.B., Silasi D.A., Steffensen K.D. & Mor G. (2008)
Cancers take their Toll—the function and regulation of Toll-like receptors
in cancer cells. Oncogene 27, 225–233.
Finn O.J. (2008) Tumor immunology top 10 list. Immunological Reviews
222, 5–8.
Ho W.Y. et al. (2003) Adoptive immunotherapy: engineering T-cell
responses as biologic weapons for tumor mass destruction. Cancer Cell 3,
431–437.
Kahn J.A. (2009) HPV vaccination for the prevention of cervical
intraepithelial neoplasia. New England Journal of Medicine 361, 271–278.
Karin M., Lawrence T. & Nizet V. (2006) Innate immunity gone awry:
linking microbial infections to chronic inflammation and cancer. Cell 124,
823–835.
Lake R.A. & Robinson B.W.S. (2005) Immunotherapy and chemotherapy: a
practical partnership. Nature Reviews Cancer 5, 397–405.
Lin W.W. & Karin M. (2007) A cytokine-mediated link between innate
immunity, inflammation, and cancer. Journal of Clinical Investigation 117,
1175–1183.
Loo D.T. & Mather JP. (2008) Antibody-based identification of cell surface
antigens: targets for cancer therapy. Current Opinion in Pharmacology 8,
627–631.
Muranski P. & Restifo N.P. (2009) Adoptive immunotherapy of cancer
+
using CD4 T cells. Current Opinion in Immunology 21, 200–208.
Murphy A. et al. (2005) Gene modification strategies to induce tumor
immunity. Immunity 22, 403–414.
Payne G. (2003) Progress in immunoconjugate cancer therapeutics. Cancer
Cell 3, 207–212.
Rafii S. (2002) Vaccination against tumor neovascularization: promise and
reality. Cancer Cell 2, 429–431.
Rabinovich G.A., Gabrilovich D. & Sotomayor E.M. (2007)
Immunosuppressive strategies that are mediated by tumor cells. Annual
Review of Immunology 25, 267 – 296.
Reichert J.M. & Valge-Archer, V.E. (2007) Development trends for
monoclonal antibody cancer therapeutics. Nature Reviews Drug Discovery
6, 349–356.
Rosenberg S.A., Yang J.C. & Restifo N.P. (2004) Cancer immunotherapy:
moving beyond current vaccines. Nature Medicine 10, 909–915.
Ruoslahti E. & Rajotte D. (2000) An address system in the vasculature of
normal tissues and tumors. Annual Review of Immunology 18, 813 – 827.
Smyth M.J. Godfrey D.I. &Trapani J.A. (2001) A fresh look at tumor
immunosurveillance and immunotherapy. Nature Immunology 2, 293 – 299.
Srivastava P.K. (2000) Immunotherapy of human cancer: lessons from
mice. Nature Immunology 1, 363–366.
Zou W. (2005) Immunosuppressive networks in the tumor environment and
their therapeutic relevance. Nature Reviews Cancer 5, 263 – 274.
Now visit www.roitt.com to test yourself on this chapter.
CHAPTER 18
Autoimmune diseases
Key Topics
The spectrum of autoimmune disease
What causes autoimmune disease
Mechanisms in autoimmune disease
Pathogenic effects of humoral autoantibody
Pathogenic effects of complexes with autoantigens
T-cell-mediated hypersensitivity as a pathogenic factor in autoimmune
disease
Some other diseases with autoimmune activity
Measurement of autoantibodies
Therapeutic options
Just to Recap...
Although most immune responses are beneficial in that they recruit
antibodies, complement, phagocytic cells, lymphocytes and so forth in
order to eliminate infectious pathogens, sometimes immunity is
inadvertently directed against antigens that do not pose a threat. The word
hypersensitivity is often used to describe such reactions. They include tissue
damaging responses to what should be innocuous environmental antigens,
as seen in allergy, and the rejection of foreign tissue introduced into the
body by the procedure of transplantation. It is clear that the diversity-
generating mechanisms involved in recombination of the V(D)J antigen
receptor genes of lymphocytes have the potential to give rise to specific
recognition of almost any antigen. One downside of this flexibility is that
some of the antigen receptors that are produced are able to recognize our
own body components—self antigens. Usually autoreactive cells that are
potentially pathogenic (e.g. those with high affinity receptors) are either
“weeded out” by the central and peripheral tolerance mechanisms, resulting
in clonal deletion and clonal anergy, or restrained by regulatory T-cells.
Introduction
In all individuals there is a degree of recognition of self. Indeed, T-cells are
required to be positively selected in the thymus for recognition of self
MHC (p. 291). Furthermore self-reactive B-cells, and self peptide + self
MHC reactive T-cells, are detectable in the circulation of all individuals, as
are autoantibodies (i.e. antibodies capable of reacting with self
components). In people without autoimmune disease the latter are
predominantly low affinity IgM autoantibodies, often produced by CD5+
B-1 cells as part of the “natural” antibody spectrum (p. 298). The term
autoimmune disease is applied when autoimmunity results in pathology.
Nonpathological autoimmunity may in fact assist in the removal of worn-
out or damaged cells and molecules. Thus, a low level of autoimmunity
seems to be the norm and generally does not result in pathology. However,
if immunological tolerance fails to eliminate or control pathogenic self-
reactive lymphocytes then autoimmune disease arises.
The spectrum of autoimmune disease
A substantial minority of individuals, estimated to be 5–8% of the
population, do however develop autoimmune disease. Once they occur,
most of these diseases then remain for life. Whilst some are relatively mild
in nature, quite a few are associated with significant morbidity and
mortality.
It is not in fact always clear whether a particular clinical entity is in fact
an “autoimmune disease” or a disease whose prime cause is not an
autoimmune attack but which, nonetheless, is associated with autoimmune
phenomenon (Table 18.1). We will discuss some such diseases, including
psoriasis and atherosclerosis, later in this chapter. There are also a number
of autoinflammatory diseases such as the hereditary periodic fever
syndromes (p. 372) characterized by an absence of high-titer autoantibodies
or autoantigen specific T-cells. Such conditions are due to malfunction of
innate immune system components and therefore do not depend upon the
breakdown of the specific immunological tolerance, which is so closely
involved in the classical autoimmune diseases.
Table 18.1 Classification criteria for autoimmune diseases. Not all these
criteria will necessarily need to be fulfilled, as clearly if will often not
be possible to demonstrate transfer of disease with autoreactive serum
and/or autoreactive lymphocytes in humans.
Indications that a disease is autoimmune
Presence of high titer autoantibodies and/or autoreactive lymphocytes in vivo
Autoantibody binding and/or T-cell reactivity to autoantigen in vitro
Transfer of disease with autoreactive serum and/or autoreactive lymphocytes
Immunopathology consistent with autoimmune-mediated processes
Beneficial effect of immunosuppressive interventions
Exclusion of other possible causes of disease
MHC association
Animal model mirroring the human disease
In the conventional autoimmune diseases the tissue distribution of the
autoantigen to a large extent determines whether the disease is “organ-
specific” or “nonorgan-specific.” Hashimoto’s disease is an example
where the antigens that are recognized are pretty much restricted to a single
organ, in this case the thyroid (Figure 18.1a). There is a specific lesion in
this endocrine gland involving infiltration by mono-nuclear cells
(lymphocytes, macrophages and plasma cells), destruction of thyroid
epithelial cells, and germinal center formation accompanied by the
production of circulating antibodies that are specific for thyroid antigens
(Milestone 18.1). In some other disorders, however, the lesion tends to be
localized to a single organ even though the antibodies are nonorgan-
specific. A good example would be primary biliary cirrhosis where the
small bile ductule is the main target of inflammatory cell infiltration but the
serum antibodies present—mainly mitochondrial—are not liver-specific.
Figure 18.1. Fluorescent antibody studies in autoimmune diseases.
(a) Thyroid peroxidase antibodies staining cytoplasm of thyroid epithelial
cells. (b) Diffuse nuclear staining on a thyroid section obtained with
nucleoprotein antibodies from a patient with systemic lupus erythematosus.
(Kindly provided by F. Bottazzo.)
Milestone 18.1—The Discovery of Thyroid
Autoimmunity
Over a century ago Sergei Metalnikoff, in 1900, reported that some male animals were able
to produce antibodies that recognized their own spermatozoa. However, these antibodies
were not pathogenic and the view of the highly respected Paul Ehrlich that the body would
not produce harmful anti-self immune responses (a situation he referred to as “horror
autotoxicus”) was at that time widely accepted. However, reports followed of self immunity
to erythrocytes (Donath and Landsteiner, 1904) and lens (Krusius, 1910). In the early 1930s
Thomas Rivers and his colleagues developed the experimental allergic encephalomyelitis
(EAE) model and provided evidence that immune cells can attack the brain. Nonetheless,
during the first half of the twentieth century there was a general air of skepticism regarding
the idea that disease could arise as a result of autoimmunity. However, during the 1940s
more reports of what seemed to be autoimmune pathology were published. Eventually any
remaining skeptics were won over in 1956 when, remarkably, three major papers from the
far corners of the globe established a link between autoimmunity and pathology in the
thyroid.
Noel Rose and Ernest Witebsky in Boston [USA] immunized rabbits with rabbit thyroid
extract in complete Freund’s adjuvant. To what one might hazard was Witebsky’s dismay
and Rose’s delight, this procedure resulted in the production of thyroid autoantibodies and
chronic inflammatory destruction of the thyroid gland architecture (Figure M18.1.1a,b).
Having noted the fall in serum γ-globulin that followed removal of the goiter in Hashimoto’s
thyroiditis and the similarity of the histology (Figure M18.1.1c) to that of Rose and
Witebsky’s rabbits, Ivan Roitt, Deborah Doniach and Peter Campbell in London (UK) tested
the hypothesis that the plasma cells in the gland might be making an autoantibody to a
thyroid component, so causing the tissue damage and chronic inflammatory response. Sure
enough, the sera of the first patients tested had precipitating antibodies to an autoantigen in
normal thyroid extracts that was soon identified as thyroglobulin (Figure M18.1.2).
The words you are searching are inside this book. To get more targeted content, please make full-text search by clicking here.
Discover the best professional documents and content resources in AnyFlip Document Base.
Search
Roitt's Essential Immunology
- 1 - 50
- 51 - 100
- 101 - 150
- 151 - 200
- 201 - 250
- 251 - 300
- 301 - 350
- 351 - 400
- 401 - 450
- 451 - 500
- 501 - 550
- 551 - 600
- 601 - 650
- 651 - 700
- 701 - 750
- 751 - 800
- 801 - 850
- 851 - 900
- 901 - 950
- 951 - 1000
- 1001 - 1050
- 1051 - 1100
- 1101 - 1150
- 1151 - 1200
- 1201 - 1246
Pages: